Sometimes there are overshots between the main slope and the beginning of it from the straigth on axis. Sometimes I find it is not easy then to know where it starts at 1 db or more to take a full octave with a book slope...12 db octave for illustration.
In my experience, it's not that simple. In a steady-state/theoretical system a statement like that may be all that's required, but in a speaker it doesn't tell you about energy storage, ringing after the impulse, etc. You can have a speaker that measures well on a standard frequency response plot that also has a hashy, nasty, ringing impulse response plot. A driver with a different cone/surround/etc. may measure similarly on frequency response and have a clean impulse plot. Small metal full-range drivers in particular seem to have more than their fair share of this behavior. It may not sound as bad as it looks, but that's a different conversation.
Distortion can also be radically different between two drivers with similar frequency response.
Focussing on slopes for the tweeter choice : should I matter for dome material or not to choose the lowest Fs in my budget in order to reduce the passive slope filter, with crossed fingers it will reduxce the number of expensive passive parts as well ?
Lowest Fs in tweeter choice = more liberty to reduce the electical then acoustical final slope ?
Lowest Fs in tweeter choice = more liberty to reduce the electical then acoustical final slope ?
Holy moly, I thought I was bad for indecision!
Why not do some sims for your intended design (XO, slope, SPL) with each candidate? Then make a list, note the price and material, rank them by ease of filter design. Sounds like you consider hard material and less filter parts needed to be good. Rank accordingly. Buy something. Try it and if you don't like it, just sell them and try something else.
Why not do some sims for your intended design (XO, slope, SPL) with each candidate? Then make a list, note the price and material, rank them by ease of filter design. Sounds like you consider hard material and less filter parts needed to be good. Rank accordingly. Buy something. Try it and if you don't like it, just sell them and try something else.
No that is not how it works: similar SPL, similar impulse. If one is clean, but the other device is ringing, a trained eye quickly sees in an unsmoothed SPL plot of a measurement with decent resolution where things go wrong with the ringing one.
I tend not to assume.
Weight was the observation.
Even if the same.
Sure the engineers are more aware of actual benefits or tradeoffs.
We have neos
Assume the material difference not all about weight.
Besides the basic obvious reasons for needing lightweight
material.
Usually going after sound quality.
Possible even same Quality, could be more with
materials and equipment and chemicals involved.
to make life easier. Or if the process is more involved
then quality becomes more obvious.
Weight was the observation.
Even if the same.
Sure the engineers are more aware of actual benefits or tradeoffs.
We have neos
Assume the material difference not all about weight.
Besides the basic obvious reasons for needing lightweight
material.
Usually going after sound quality.
Possible even same Quality, could be more with
materials and equipment and chemicals involved.
to make life easier. Or if the process is more involved
then quality becomes more obvious.
diyiggy: it is not that simple to say certain tweeter will work with low order filter well. Good way to find out and learn is to study the projects others did. For example see my Jurko project, TW29RN with just resistor (level matching to midwoofer) and capacitor plus RLC to tame Fs peak. From my experience, this is likely the simplest real world topology that works.
https://pkaudio.webnode.cz/jurko/
Attachments
Last edited:
It doesn't work so simply I'm afraid. Because you have to account for the frequency-varying impedance of the driver -particularly, if you have a particular desire for one reason or another, the peak at Fs, which will foul up the transfer function and also reduce power-handling. One of the foibles of low order acoustical slopes is that you can often end up needing more, not fewer, components than a higher order filter because you are likely to run into these issues & need to correct for them.
Possibly, but it isn't a given -it depens as much on FR, impedance & power-handling (electrical inc. thermal & available low-distortion excursion). You can't really isolate one factor like that, unfortunately.
For the most part.
Speaker is what speaker is. real world is listening.
Bur generalized basic impedance curve is gonna tell yah
what any driver is likely to do and not do well.
rather a sub or a tweeter, eventually the piston or surround gives
up the ghost. not rocket surgery.
higher order filters can help reduce how soon it gives up the ghost.
and just the impedance curve and the peaks will tell you what any speaker
doesn't want do anymore. its over at that point. rather its 30 dollars
or 300 dollars.
And if you trying to weasel as much as you can towards that peak.
gonna need a high order filter for high SPL
Speaker is what speaker is. real world is listening.
Bur generalized basic impedance curve is gonna tell yah
what any driver is likely to do and not do well.
rather a sub or a tweeter, eventually the piston or surround gives
up the ghost. not rocket surgery.
higher order filters can help reduce how soon it gives up the ghost.
and just the impedance curve and the peaks will tell you what any speaker
doesn't want do anymore. its over at that point. rather its 30 dollars
or 300 dollars.
And if you trying to weasel as much as you can towards that peak.
gonna need a high order filter for high SPL
How do you know when the Fs peak of a tweeter should be tammed, please ? When too much high (> to 15 ohms for instance ?) or it should be just tested as a try and error with listening checking in every designs ? In Dickason book I remember having read it is not worthing it when the tweeter is alreday ferro fluided ?!
@Pida , hello, I like your designs a lot, I do not know how your wood worker does to take such beautifull front baffles with smooth edges. 🙂
@Pida , hello, I like your designs a lot, I do not know how your wood worker does to take such beautifull front baffles with smooth edges. 🙂
Last edited:
A bicycle analogy comes to mind. "Al 6061 aircraft grade alloy blah blah" are advertised as being lightweight, but I think the reality is lower costs, and welding at a mere 700-or-so degrees C is probably lighter on the power bill. Yet the resulting bikes still weigh exactly the same, if not more than the "low tech" steel bikes that are now unobtainium.
Because magnesium is lighter, they can use a thicker layer to stiffen the dome.
By the way, I need to dig up that short video with a converging wave simulator in a circular pool.
Because magnesium is lighter, they can use a thicker layer to stiffen the dome.
By the way, I need to dig up that short video with a converging wave simulator in a circular pool.
every situation is different. but I tend to accept regardless.
If you are crossing close to FS usually because of a larger mid.
Low SPL levels it can work, just higher levels its gonna distort quicker.
Otherwise look at Fs double it, 3rd order or 4th
all done. eliminates most the surround/excursion distortion.
any or most distortion gonna be the cone shape itself or material
at high levels.
Never always a perfect world though. most the time ill try to weasel
as close as possible with 3rd order to go lower. still steep filter helps a lot.
2nd orders easier you can do it. just not gonna be surprised, it works well
at low SPL only. that is the tradeoff. 2nd makes it easier.
but yeah double Fs min. its been done otherwise.
distortion comes early with SPL its that simple
If you are crossing close to FS usually because of a larger mid.
Low SPL levels it can work, just higher levels its gonna distort quicker.
Otherwise look at Fs double it, 3rd order or 4th
all done. eliminates most the surround/excursion distortion.
any or most distortion gonna be the cone shape itself or material
at high levels.
Never always a perfect world though. most the time ill try to weasel
as close as possible with 3rd order to go lower. still steep filter helps a lot.
2nd orders easier you can do it. just not gonna be surprised, it works well
at low SPL only. that is the tradeoff. 2nd makes it easier.
but yeah double Fs min. its been done otherwise.
distortion comes early with SPL its that simple
WhiteDragon is correct - we don't all listen to tweeters only at low SPL.
There is even a 'general rule' for cross-over design, and that is:
For a first order - your C/O frequency should be at least 2 octaves above Fs.
For a second order - your C/O frequency should be at least 1 octave above Fs.
This in itself generally tames any normal lower frequency impedance peak - sonically.
This 'rule' then generally dictates a 3way design, rather than 2way. Ferro-fluid can really help 2way design.
I only rarely use third order design - for special situations IE. 'strange' drivers and very high SPL.
BACK TO DIAPHRAGM MATERIALS
Just like metal cone woofers, most good metal diaphragms no longer use just pure metal, but also incorporate
extra materials to tame 'Ring & Resonance'. This makes good sense to me.
There is even a 'general rule' for cross-over design, and that is:
For a first order - your C/O frequency should be at least 2 octaves above Fs.
For a second order - your C/O frequency should be at least 1 octave above Fs.
This in itself generally tames any normal lower frequency impedance peak - sonically.
This 'rule' then generally dictates a 3way design, rather than 2way. Ferro-fluid can really help 2way design.
I only rarely use third order design - for special situations IE. 'strange' drivers and very high SPL.
BACK TO DIAPHRAGM MATERIALS
Just like metal cone woofers, most good metal diaphragms no longer use just pure metal, but also incorporate
extra materials to tame 'Ring & Resonance'. This makes good sense to me.
what are these good metal dome tweeters you talk about, please ? I am aware of such mix in PA compression drivers, but not so for hifi dome tweeters.
Those lasts most of the time just blend magnesium with aluminum and the extra material is just a coating to avoid corosion (especially with Magnesium) : ceramic oxide for instance in the SB26CDC, shelac or gold (Audax) or diamond deposit on some others. The only one I remember to be not pure metal was the Eton 2 tweeter which was ceramic on two sides on a core foam but that's not metal. I am talking of course of the ones diy enthusiasts can source.
The internal damping and ringing breakup point seems to stay the atribute of the core material. (Be is beyond my budget)
Those lasts most of the time just blend magnesium with aluminum and the extra material is just a coating to avoid corosion (especially with Magnesium) : ceramic oxide for instance in the SB26CDC, shelac or gold (Audax) or diamond deposit on some others. The only one I remember to be not pure metal was the Eton 2 tweeter which was ceramic on two sides on a core foam but that's not metal. I am talking of course of the ones diy enthusiasts can source.
The internal damping and ringing breakup point seems to stay the atribute of the core material. (Be is beyond my budget)
Most domes only have a single point (or 'ring') where most of the damping occurs and whatever forces the voice coil applies are always shared between the dome and the surround. By the same token, whatever braking forces the surround exerts, they have to get past the VC + amplifier's EMF first in order to be able to dampen the dome. So the amplifier + whatever passive filters are critical here. The physical design 'recipe' for a dome tweeter is not ideal, IMO, but it is what it is.
AFAICT, it shouldn't be necessary for the dome itself to be mechanically damped, as long as the electrical design doesn't interfere. So that probably means adding an RLC notch to disconnect the amplifier at some break-up frequency.
AFAICT, it shouldn't be necessary for the dome itself to be mechanically damped, as long as the electrical design doesn't interfere. So that probably means adding an RLC notch to disconnect the amplifier at some break-up frequency.
Do you maybe mean low passing the tweeter? There has been talk of applying a high impedance at the breakup frequency but it's in relation to modifying the behaviour of harmonic distortion in a low passed driver.
Otherwise you can't stop the breakup electrically, but you can smooth the response on an axis.